*3.2. CRISPR–Cas9-Mediated Disruption of Endogenous HSP23 and HSP100 Genes in L. braziliensis*

Next, we tested the applicability of the PCR-based CRISPR–Cas9 method on two endogenous, single-copy genes of *L. braziliensis* encoding the heat shock proteins HSP23 and HSP100. Both genes were successfully replaced in Old World *Leishmania* spp, using homologous recombination, giving rise to conditional phenotypes [47,56,57]. Previous work in *L. donovani* showed that *HSP23* null mutants are sensitive to temperature and chemical stresses. In *L. major* and *L. donovani*, ∆*clpB* (*HSP100*) null mutants showed loss of virulence in vitro and in vivo. We sought to replicate those findings in *L. braziliensis* to assess the practical application of CRISPR–Cas9-mediated genetic manipulation in this parasite species. First, we tested the fitness of *L. braziliensis* (Cas9/T7) cells by *in vitro* growth analysis (Figure S1C) and found slightly increased proliferation compared with wild type cells, thus excluding overt, detrimental effects of Cas9 expression.

For disruption of each targeted GOI, the *L. braziliensis* Cas9/T7 parental cell line was transfected in parallel with four different sets of sgRNAs and donor DNAs (see Table S1 for nucleotide sequences). Double drug-resistant cell populations for both targeted genes emerged in culture at day 18 post transfection, and were then subjected to a higher drug selection pressure, as established for *eGFP* deletion.

#### 3.2.1. *LbrHSP23* Gene Replacement

Three pairs of sgRNAs targeted different sites within the *LbrHSP23* ORF (Figure 2A), while a fourth pair of sgRNAs was designed to create DSBs upstream and downstream of the GOI coding region for whole-gene deletion (not shown). Putative *HSP23-*null mutants were obtained with sgRNAs sets 1 and 2 (Figure 2A), both of which disrupted the alpha-crystallin domain of HSP23, a conserved signature feature of the small heat shock protein family [58]. Transfection with sgRNAs set 3, which targeted the C terminal part of LbrHSP23, did not generate viable cells after double selection. No *LbrHSP23* whole-gene deletion mutants could be obtained with sgRNAs set 4, either. Later analysis revealed a one-base pair mismatch between primer P4-LbrHsp23–3'sgRNA (Table S1) and the *L. braziliensis* strain PER005 *HSP23* gene, explaining the lack of success for sgRNA set 4.

From the transfections with sgRNAs sets 1 and 2, three cell populations emerged: one with set 1 at day 18 post-transfection, and two with set 2, at day 18 and 25 post-transfection, respectively. From these three populations, clones were raised and expanded. Three clones were then subjected to whole genome sequencing: *HSP23*–/– cl.1 and cl.2, from transfection with sgRNAs set 2; and *HSP23*–/– cl.3, derived from the transfection with sgRNAs set 1. NGS analysis verified a lack of sequence reads for the targeted gene regions (Figure 2B), confirming site-specific disruption of the *LbrHSP23* ORF. Moreover, the precise integration of both drug-resistance cassettes in these *HSP23*–/– mutant clones was also verified (Figure S6). Western blot analysis using specific antibodies [47] failed to detect HSP23 protein in the *HSP23*–/– mutants (Figure 2C), confirming the null mutants on the genomic and proteomic levels.

**Figure 2.** CRISPR–Cas9-mediated disruption of the endogenous *HSP23* gene in *L. braziliensis*. (**A**) Schematic representation of the *LbrHSP23* locus depicting the locations of 20-nt guide sequences that worked efficiently to disrupt the *LbrHSP23* ORF. Two sets of sgRNAs were tested (set 1 and set 2): set 1 = *LbrHSP23*-70-5'sgRNA and *LbrHSP23*-171-3'sgRNA; set 2 = *LbrHSP23*-183-5'sgRNA and *LbrHSP23*-323-3'sgRNA. Both pairs are designed to disrupt the conserved functional alpha-crystallin domain of HSP23 (amino acid positions 6–104). The guide sequence pairs with the DNA target (blue bar) directly upstream of a requisite 5′–NGG–3′ adjacent motif (PAM). The green arrowhead indicates the predicted Cas9 cleavage sites. Only the coding strand sequence is shown. (**B**) NGS analysis of the *HSP23* locus after CRISPR–Cas9-mediated gene replacement. Genomic DNA of *L. braziliensis* PER005cl2 wild-type parasites (WT), the parental cell line WT [Cas9] and *HSP23*–/– mutant clones was isolated and subjected to NGS analysis. Resulting NGS reads were aligned to the *HSP23* gene locus (LbrM.20.0220) in the *L. braziliensis* M2904 reference genome using the Bowtie 2 algorithm. The read coverages (Y-axis) for the gene locus are shown in blue. The arrow represents the position and direction of the coding sequence. The X-axis numbering refers to the nucleotide position (bp) on chromosome 20. Grey-shaded areas denote lack of aligned reads. (**C**) Verification of *HSP23* gene replacement by Western blot analysis. 1 × 10<sup>7</sup> cells of WT, WT [Cas9], and of 3 *HSP23*–/– clones were lysed and the cell lysates were analysed by SDS-PAGE and Western blot using anti-HSP23 (1/500, lower panel). Anti-HSP100 (1/1000, upper panel) was used as loading control. MW = Molecular weight in kilodalton.
